Mike @MikeLongTerm

$AMD shareholders gonna feel it when we hit 50%+ revenue share before year end and 50%+ market share by next year

Patrick Moorhead @PatrickMoorhead

6 hours ago

AMD, in 9 years, went from 0% server revenue share to 46%. What a feat. techpowerup.com/349029/amd-epy…

6 19 143 8K 9

$AMD shareholders just realized 1 tweet would be more than enough to take @AMD to new ATH than 5,000 $AMD long threads @MikeLongTerm wrote.

Mike @MikeLongTerm

11 hours ago

$AMD| The FOMO to buy @AMD Chips is NOW 🧵 Not Financial Advice! DYOR! Research Purpose Only! The Inference Queen is the biggest winner in Agentic AI where all other CPUs are struggling to compete with a 2yr old EPYC Turin and EPYC Venice is in mass production phase. AMD

1 9 36 23K 7

Cash Secure Put on $META & $MSFT and posted for subscribers. Not Financial Advice!DYOR!

Mike @MikeLongTerm

6 hours ago

$PLTR $AMD $TSLA $HIMS $GRAB $XYZ RIP BEARS I thought they said the end of the world this week It is hard to be a bear these days.

2 1 14 5K 0

$PLTR $AMD $TSLA $HIMS $GRAB $XYZ RIP BEARS I thought they said the end of the world this week It is hard to be a bear these days.

zerohedge @zerohedge

6 hours ago

Trump says he has "cancelled the scheduled strikes and bombings against Iran this evening" as "discussions with the Islamic Republic of Iran have been brought to the highest level of Iranian leadership and approved"

199 141 1K 195K 53

$HIMS| What do you see?

Mike @MikeLongTerm

2 weeks ago

$HIMS $100 stock that Investors are willing to lend at 0% interest| Convertible Notes 🧵 Not Financial Advice! DYOR! Valuation is ridiculous at this kind of growth(60-70% FY2026 with Eucalyptus acquisition) P/S : 2.34x Fwd P/S: 1.37x This is clearly pricing in the

7 10 130 107K 42

OpenAI had most of market share on B2C for years, until $GOOGL makes Google Gemini generous on free tier. If there is no competition, Enterprises would be forced to use it even if it is this expensive. But that is not the reality. As far as valuation, 21x P/S FY2026 is very reasonable for a young company like Anthropic.

$PLTR is 60%+ Adjusted Operating margin @AnthropicAI is 5% Adjusted Operating margin There is no such thing as unlimited budget for Tokens. The cost has to make sense economically. Enterprises are not stupid. If you only there to sell tokens, that is not a good long term strategy. And bills rised this much and only 5% Adj Operating Margin, may be you have a renting compute too much problem. Just may be, you need to own the data center to lower the cost for customers.

Mike @MikeLongTerm

a day ago

BREAKING $PLTR @PalantirTech Dr. Karp Brandnew Interview on CNBC 🚀🚀🚀 Software companies are trying too hard to copy Palantir (FDEs) but failed to beat Rule of 40 at ATH 145%! "Do you understanding AI?" - Alex Karp Bullish on @elonmusk running @SpaceX Space Business. No

13 34 300 67K 172

$AMD @tensorwave 2 GW x.com/MikeLongTerm/s…

Mike @MikeLongTerm

2 days ago

BREAKING $AMD & @tensorwave Muli-GW 👀👀👀 In the artificial-intelligence world, pretty much everyone wants to lay hands on chips and other computing hardware made by the market leader, Nvidia. Everyone except Darrick Horton, that is. Horton is the 28-year-old co-founder and

3 7 45 10K 3

$AMD Agentic AI Rack At SCALE is out! x.com/MikeLongTerm/s…

Mike @MikeLongTerm

2 days ago

BREAKING $AMD Agentic AI Rack At SCALE is out! 🚀 AKA CPUs Dense Rack for Fleets of Agents!🆕🆕🆕 I been talking about this for years & it is HAPPENING Not Financial Advice! DYOR! Today, @AMD published a detailed technical blog emphasizing that the future of agentic AI

3 11 69 24K 9

$AMD @tensorwave long-term partnership & how to scale to 2GW to help bring down compute cost and improve accessibility. Will be interesting if @tensorwave is going to one of the first to adopt Agentic AI Rack(EPYC dense) to bring down Inference cost to accelerate further adoption on ROCm. The AMD-Native Bet (0:52 - 2:03): Unlike the majority of NeoClouds focused on Nvidia, TensorWave went all-in on AMD due to their prior experience with FPGAs and confidence in AMD's open-source roadmap. This specialized focus allows them to provide a robust alternative in a market facing severe supply constraints. Bridging the Software Gap (2:31 - 4:14): Addressing concerns about the CUDA ecosystem, Tatarchuk highlights that AMD has made significant progress in software reliability. TensorWave has further simplified this by developing their own open-source training and inference stack, Scaler LM, created by original members of the Nvidia team who built CUDA, allowing for easier transitions between GPU architectures. Series B Milestone (4:25 - 4:40): The company recently closed a $350 million Series B funding round led by AMD and Magnetar, valuing the company at $1.55 billion. This capital will be used to secure data center capacity and scale infrastructure to meet the massive demand for AI compute. Addressing Bottlenecks (6:21 - 6:54): While GPU availability is improving, Tatarchuk identifies access to power and data center capacity as the current primary bottleneck. TensorWave is focused on locking up this infrastructure for 2026 and 2027 to ensure they can scale alongside the surging demand from enterprise and AI labs. Las Vegas as a Tech Hub (15:02 - 16:50): TensorWave is headquartered in Las Vegas, which Tatarchuk describes as an untapped tech zone. He cites tax advantages, accessibility for clients, and a business-friendly environment as key drivers for their growth in Nevada, supported by local state investment. Future Outlook (17:03 - 17:27) As they look ahead, TensorWave's top priorities include aggressive team expansion, acquiring more GPUs, and securing significantly more data center capacity to meet their goal of moving into gigawatt-scale infrastructure by 2027. Source & Credit: youtube.com/watch?v=uKWTHh…

Mike @MikeLongTerm

11 hours ago

$AMD| The FOMO to buy @AMD Chips is NOW 🧵 Not Financial Advice! DYOR! Research Purpose Only! The Inference Queen is the biggest winner in Agentic AI where all other CPUs are struggling to compete with a 2yr old EPYC Turin and EPYC Venice is in mass production phase. AMD

1 9 36 23K 7

if you are still missing the point, this isn't abt compute anymore, when cost rised this much and this fast. Enterprises complained for a reason. Tokenomic is now the center of the discussion. Helios Rack at scale like 1st GW for OpenAI can bring cost down to $0.0003-$0.0005/M Tokens, that will mean Anthropic will be forced to own AMD chips and compete. Competition will bring token cost down, which will not grow revenue like it is now. Companies are already capping token budget per month basis.

False. Shortage of compute does not mean enterprises have an unlimited budget for tokens where cost cannot justify the productivity gain. In the last 3 months, since the rising adoption of Agentic AI, companies came on CNBC/Bloomberg to complain about Token costs/bills. This is not sustainable. New compute will come online, but TCO is much higher from $NVDA chips, which does not lower token cost on Inference. Only $AMD chips can bring down inference cost meaningfully, and AMD has to wait for TSMC to ramp. People/Enterprises are canceling Anthropic due to cost x5-20x what it was budgeted for. And also false, OpenAI frontier models are very competitive in 2026 now they getting Helios Rack, which allow OpenAI to be more token cost competitive. Revenue base is much higher by end 2026, where most wont see that kind of growth in 2027.

@schartz_c @AMD Have you read the thread below on Token cost difference? Lower cost will increase consumption by 10x-100x, and allow companies to gain more productivity at lower cost. And yes token cost from renting $NVDA chips or Hyperscalers is very high right now for Anthropic

$AMD shareholders and analysts just realized only @AMD chips could solve the Tokenomic Crisis where Dr. Su been talking abt this for nearly 4 years. Token cost didnt matter until Agentic AI where 40-60m CPUs units highest core/thread matters to bring down $/M tokens.

Mike @MikeLongTerm

11 hours ago

$AMD| The FOMO to buy @AMD Chips is NOW 🧵 Not Financial Advice! DYOR! Research Purpose Only! The Inference Queen is the biggest winner in Agentic AI where all other CPUs are struggling to compete with a 2yr old EPYC Turin and EPYC Venice is in mass production phase. AMD

1 9 36 23K 7

@Interestingna11 @AnthropicAI No. Enterprises are canceling Anthropic due to rising cost. Growth will look good at IPO, but slow down after. They purposely built model significant more tokens to jack up the revenue. @OpenAI has more leverage with lower token cost. But this post isnt abt valuation.

The amount of AI generated ppt slides customers share with me on their projects/architecture and how much they all look the same now. This should tell you something of how much vibe coding will just create variations of the same thing of marginal value.

Selling 15-30 forward P/E to buy 80x forward P/S. And retail investors expect to 10x money in 2 hours

Polymarket @Polymarket

10 hours ago

JUST IN: SpaceX IPO has reportedly drawn $70,000,000,000.00 in orders from retail investors.

304 262 4K 475K 205

$GRAB Dine-Out Expansion 🚀🚀🚀 Grab Dine Out (also called Dine Out Deals or similar) is Grab's feature in its super app for discovering, booking tables at, and getting discounts on dine-in restaurants across Southeast Asia (Singapore, Malaysia, Indonesia, Thailand, Philippines, Vietnam...). Users browse options in the app, book tables, and often get instant discounts (up to 20-50% off the bill, sometimes stacked with promos like "Dine Out Tuesdays" or extra codes) when paying via Grab SEA consumers increasingly use apps like Grab to explore new restaurants, cuisines, and spots before deciding to go out. Time spent browsing on the platform has risen (~17 minutes per session), with users researching from home or even during rides. Grab added tools like the 5-Star Food Merchant Program to highlight top spots. This is one of new service that has a very high margin within the SuperApp Portfolio with every low acquisition cost from customers and merchants. According to Grab’s consumer insights, 84% of surveyed users said they had used the Grab app to discover new restaurants or stores, while 61% said they often browse without a specific merchant in mind. While this presents an opportunity for eateries to reach customers at the point of discovery, standing out remains a challenge. Under the initiative, diners can look forward to neighbourhood-specific campaigns on Grab Dine Out, developed in partnership with Enterprise Singapore. At participating independent F&B outlets, customers can enjoy offers of up to 15% off their total bill. The first precinct campaign is already taking place in Tanjong Pagar until 25 July, with more planned later this year in Holland Village and Jalan Besar. There will also be nationwide delivery deals on GrabFood, designed to help smaller and independent merchants capture demand beyond their immediate neighbourhoods. These Grab-funded promotions will allow consumers to enjoy up to 15% off menu-wide via the Food tile on the Grab app.

Mike @MikeLongTerm

3 weeks ago

BREAKING $GRAB CEO @AnthonyPY_Tan New Interview The biggest thing, where Anthony talked abt Taiwan Expansion and he wants to expand way beyond just SEA. Again, I said this last year, Bottom of Pyramid is 5 Billion people TAM, if he is sticking to this mission, $GRAB will worth

6 18 120 47K 39

June 2026 will be an interesting month, where most will look back 6 months from now and cry about how they missed it.

PPI 1.1% MoM, Exp. 0.7% PPI Core 0.4% MoM, Exp. 0.5% PPI 6.5% YoY, Exp. 6.4% PPI Core 4.9% YoY, Exp. 5.4%